Re: In-kernel units for block numbers, etc ...

[mlelstv%serpens.de@localhost (Michael van Elst)]

rhialto%falu.nl@localhost (Rhialto) writes:

>Would it make sense to create a separate layered device that converts
>the block size of its underlying device? At least that would separate
>the problem and make it simpler. And it might be useful in its own
>right, for instance when you dd a file system image to a drive with
>different blocksize.

Using a drive with small blocks to emulate a drive with large blocks
breaks atomicity of I/O operations. While that often isn't even true
when using the native block sizes due to write caching, it is still
more reliable than such an emulating driver.

Other than that, there are some practical problems with such a driver,
like seeing the same disklabel on both levels.

-- 
-- 
                                Michael van Elst
Internet: mlelstv%serpens.de@localhost
                                "A potential Snark may lurk in every tree."